JP5575648B2 - Cluster injection processing method - Google Patents

Description

  The present invention relates to a cluster injection processing method for processing a sample surface using a reactive cluster formed from a reactive gas, and a semiconductor device and a microelectromechanical device manufactured by a processing method using a reactive cluster. And an optical component.

  As a method of processing by irradiating the sample surface with the gas cluster, for example, the gas cluster is ionized, accelerated by an electric field or magnetic field, and collided with the sample surface to remove atoms and molecules on the sample surface. There is a method using an ion beam (see, for example, Patent Document 1).

  In the sample processing method described above, a gas consisting of a massive atomic group or molecular group of a gaseous substance is ejected from a gas supply unit by injecting a pressurized gas obtained by mixing a gaseous substance at normal temperature and pressure and a rare gas. Create a cluster. The gas cluster is ionized by irradiation with an electron beam to form a gas cluster ion beam.

When this gas cluster ion beam is irradiated onto the solid surface, multistage collisions occur between molecules or atomic species constituting the cluster ions and atoms on the solid surface. For this reason, a reflected atom or molecule having a lateral motion component is generated, and the substrate surface can be planarized or cleaned by the reflected atom or molecule.
In this way, the surface of the sample can be processed by the gas cluster ion beam.

  Moreover, the acceleration voltage applied to the cluster ion beam is reduced stepwise or continuously by providing the processing apparatus using cluster ions with an accelerating electric field unit for applying an accelerating voltage to the cluster ions and a decelerating electric field unit for applying a decelerating voltage. Has been proposed (see, for example, Patent Document 2).

JP-A-8-319105 Japanese Patent Laying-Open No. 2005-310977

  However, in a conventional method using a gas cluster ion beam or an apparatus using a gas cluster ion beam, a step of ionizing the cluster and accelerating ions with a DC voltage is performed. For this reason, the cluster ion is electrically changed to neutral using a weak ionized molecule or a neutralizer, and then collides with the sample.

  However, in the above-described method, the cluster cannot be made into completely neutral particles, and electrical damage to the sample cannot be completely eliminated.

Also, for example, when manufacturing semiconductor elements, micro electro mechanical systems (MEMS), optical components, etc., deep processing with an aspect ratio of 1 or more, formation of through holes in a substrate, metal layer Etching is performed. Such processing is generally performed by plasma anisotropic etching.
However, if the above-described processing is performed by plasma anisotropic etching, for example, characteristics such as a threshold voltage of a controller / transistor composed of a MOS transistor, a MOS capacitor, or the like already formed on the substrate change. In addition, for example, in a multilayer wiring structure formed in a semiconductor element, the dielectric constant of an interlayer insulating film made of a material having a low dielectric constant (Low-k) changes due to plasma heat.
In addition to the plasma treatment, for example, etching by an immersion method in an etching solution is also performed. However, etching by plasma treatment or dipping method generates a large amount of side etching or undercut, and is not suitable for miniaturized or densified patterns.

  In order to solve the above-mentioned problems, the present invention provides a sample processing method using electrically neutral reactive clusters. In addition, the present invention provides a semiconductor element, a microelectromechanical element, and an optical member that are processed by anisotropic etching by applying the above-described processing method.

Cluster jet processing method of the present invention, the gas mixture from the reaction gas consisting of low-boiling gases, a pressure in the range which does not liquefy, jetted while adiabatically expanded in a predetermined direction, an electrically neutral reactive A reactive cluster is generated, the reactive cluster is sprayed onto a sample in a vacuum processing chamber, and the surface of the sample is processed by a chemical reaction .

According to the present invention, when a sample is processed, a reactive cluster is formed using a reactive gas and a gas having a lower boiling point than the reactive gas, so that the reactive cluster collides with the sample and reacts with the sample. Thus, the sample surface can be processed. Therefore, it is not necessary to ionize the gas cluster by applying an electric field or a magnetic field, and it is not necessary to make the cluster electrically neutral because of the electrically neutral particles that are not ionized. Therefore, no electrical damage is given to the sample. Furthermore, since no sheath is generated, no reactive clusters are drawn in the side wall direction, and the straightness in the depth direction of anisotropic etching is improved.
Therefore, according to the present invention, it is possible to perform processing with high straightness in the depth direction while eliminating electrical damage to the sample.
Further, according to the present invention, the semiconductor substrate is processed by jetting the reactive clusters into the vacuum processing chamber. Therefore, highly accurate anisotropic etching can be performed without performing plasma anisotropic etching. Therefore, deep processing with an aspect ratio of 1 or more can be performed on a substrate or the like without causing a change in the threshold voltage of the transistor due to plasma treatment, a change in the dielectric constant of the interlayer insulating film, or the like. In addition, by highly accurate anisotropic etching, the amount of side etching and undercut generation is very small, and a highly dense and fine pattern can be formed. For this reason, it is possible to provide a semiconductor element, a microelectromechanical element, and an optical member with little deviation from the design value by highly accurate anisotropic etching using reactive clusters.

It is a figure for demonstrating the structure of one Embodiment of the processing apparatus of the sample concerning the processing method of this invention. In the processing method of this invention, it is a figure which shows the relationship between the primary pressure and the injection time to a sample, and the etching rate of a sample. In the processing method of this invention, it is a figure which shows the relationship between the distance of an ejection part and a sample surface, and the etching rate of the sample by processing. In the processing method of this invention, it is a photograph showing the state of the sample surface after a process at the time of processing as the distance of an ejection part and a sample surface being 26 mm. In the processing method of this invention, it is a photograph showing the state of the sample surface after a process at the time of processing as the distance of an ejection part and a sample surface being 35 mm. In the processing method of this invention, it is a photograph of the cross section near the sample surface after a process at the time of processing a sample surface by making the incident angle of a reactive cluster into the diagonal direction with respect to the sample surface. In the processing method of this invention, it is a photograph of the state of the silicon single crystal in which the through-hole was formed. The ClF ₃ gas used alone is a photograph showing a condition of the surface of the sample after processing in the case of performing processing of the sample surface. It is a photograph showing the state of the sample surface after processing when processing the sample surface using Ar gas alone. It is a photograph showing the state of the sample surface after a process at the time of processing a sample surface by making primary pressure into -0.09MPaG. Is a photograph showing a condition of the surface of the sample after processing in the case of processing using Cl ₂ gas. Is a photograph showing a condition of the surface of the sample after processing in the case of processing using the F ₂ gas. It is a photograph which shows the state of the sample surface after a process at the time of processing the sample in which the silicon oxide film was formed in the processing method of this invention. In the processing method of the present invention, a photograph showing the state of the sample surface after processing when the surface of the sample from which the natural oxide film has been removed is processed It is a photograph showing the state of the sample surface after a process at the time of processing the surface of a sample, without removing a natural oxide film in the processing method of this invention. In the processing method of this invention, it is a figure which shows the relationship between the primary pressure and the injection time to a sample, and the etching rate of a sample. It is a figure which shows schematic structure of embodiment of the semiconductor element of this invention. A to C are manufacturing process diagrams of the embodiment of the semiconductor device of the present invention. It is a figure which shows schematic structure of embodiment of the pressure sensor of this invention. It is a figure explaining the manufacturing process of embodiment of the pressure sensor of this invention. FIG. 3A is a manufacturing process diagram of an embodiment of an optical element of the present invention. B is a schematic block diagram of an embodiment of an optical element of the present invention. FIG. 3A is a manufacturing process diagram of an embodiment of an optical element of the present invention. B is a schematic block diagram of an embodiment of an optical element of the present invention.

Hereinafter, specific embodiments of the present invention will be described with reference to the drawings.
First, an embodiment of a sample processing apparatus applied to the processing method of the present invention is shown in FIG.
The processing apparatus shown in FIG. 1A includes a gas supply unit 11 and a sample stage 14 in a vacuum processing chamber 13.

The gas supply unit 11 is provided with an ejection unit 12 having an opening having a sufficiently small area for adiabatic expansion. Further, the ejection part 12 is provided in the vacuum processing chamber 13.
In addition, a gas supply unit (not shown) supplies the gas supply unit 11 with a mixed gas of an interhalogen compound or hydrogen halide, which is a reactive gas, and a gas having a lower boiling point than the reactive gas. And the molecule | numerator 18 of the mixed gas with which the gas supply part 11 was filled is ejected from the ejection part 12, and the reactive cluster 19 is formed.

  The ejection part 12 includes, for example, an opening formed in a circular shape as shown in FIG. 1B and an opening formed in a rectangular shape as shown in FIG. 1B at the end of the gas supply unit 11. In addition, the ejection part 12 may have any shape as long as it is an opening having a sufficiently small area from which a differential pressure capable of forming a reactive cluster can be obtained between the gas supply part 11 and the vacuum processing chamber 13.

In the vacuum processing chamber 13, the gas supply unit 11 and the sample stage 14 for installing the sample 17 are installed so as to be aligned on a straight line in the gas injection direction. Moreover, the sample stage 14 and the injection part 12 are movably installed by driving a movable part (not shown) in the vacuum processing chamber 13.
For this reason, by moving the sample stage 14 and the injection unit 12 or changing the inclination angle, the distance to the ejection unit 12 is changed, or the position at which the reactive cluster ejected from the ejection unit 12 collides is changed. Or the incident angle of the reactive cluster can be changed.

  A turbo molecular pump 15 and a dry pump 16 are connected to the vacuum processing chamber 13. By combining these two pumps, the inside of the vacuum processing chamber 13 can be decompressed.

  Next, a sample processing method using the processing apparatus shown in FIG. 1 will be described as an embodiment of the sample processing method of the present invention.

First, as the reactive gas, for example, a mixed gas of an interhalogen compound or hydrogen halide and a gas having a lower boiling point than the reactive gas is supplied to the gas supply unit 11 using a supply unit (not shown). The pressure (primary pressure) in the gas supply unit 11 at this time is a pressure within a range where the reactive gas having a high boiling point is not liquefied, for example, 0.3 MPaG in the case of a mixed gas of 6 vol% ClF ₃ and 94 vol% Ar. More than 1.0 MPaG.
The primary pressure uses PaG, which is a pressure unit based on atmospheric pressure, and the secondary pressure uses Pa, which is a pressure unit based on absolute pressure.

Then, the inside of the vacuum processing chamber 13 is evacuated by the turbo molecular pump 15 and the dry pump 16, and the pressure (secondary pressure) in the vacuum processing chamber 13 is controlled to 100 to 0.1 Pa or less, particularly preferably 10 Pa or less. To do. Due to the difference between the primary pressure of the mixed gas in the gas supply unit 11 and the secondary pressure in the vacuum processing chamber 13, the mixed gas in the gas supply unit 11 is injected from the ejection unit 12 into the vacuum processing chamber 13. .
At this time, the gas injection from the ejection part 12 is performed by the differential pressure between the primary pressure and the secondary pressure. For this reason, it is desirable to increase the difference between the primary pressure and the secondary pressure. However, since the gas cannot be stably supplied in a state where the mixed gas is liquefied, it is necessary to perform the primary pressure below a pressure that does not liquefy the mixed gas.
Further, the kinetic energy when the pressure difference between the primary pressure and the secondary pressure is larger is ejected from the ejection portion 12. For this reason, the energy at the time of a reactive cluster colliding with a sample becomes large, and the reaction energy which processes a sample becomes large.

  At this time, the injected mixed gas suddenly changes from a high pressure state to a low pressure state, and an electrically neutral reactive cluster is generated. The reactive cluster is usually formed from a group of atoms or molecules constituting tens to thousands of reactive gases, such as interhalogen compounds or hydrogen halides, and a gas having a lower boiling point than the reactive gas.

The reactive gas needs to have high reactivity with the sample to be processed. For this reason, when using a silicon single crystal as a sample, it is preferable to use an interhalogen compound having high reactivity with silicon. As interhalogen compounds, for example, ClF, ClF ₃ , ClF ₅ , BrF ₃ , BrCl, IF ₅ , IF ₇ can be used. In particular, it is preferable to use ClF ₃ having high reactivity with silicon.
In addition, when a metal material or a semiconductor material other than silicon, for example, GaAs, InP, GaN, or the like is used as a sample, it is preferable to use hydrogen halide having high reactivity with the metal material.
As the hydrogen halide, HCl, HBr, HI or the like can be used. In particular, it is preferable to use HI having high reactivity with a metal material.

At this time, since the reactive gas described above has a high boiling point, when only the reactive gas is supplied into the gas supply unit 11, the pressure necessary for generating the reactive cluster cannot be obtained. For this reason, the partial pressure of the above-mentioned reactive gas can be reduced by mixing the above-mentioned reactive gas, for example, a gas having a low boiling point with respect to the interhalogen compound and hydrogen halide. For this reason, primary pressure sufficient to produce a reactive cluster can be obtained while preventing liquefaction of the reactive gas.
Examples of gases having lower boiling points than reactive gases include rare gases such as He, Ar, Ne, Kr, and Xe, N
_{2, CO 2, O 2} or the like can be used.
Further, it is desirable that a gas having a lower boiling point than the reactive gas does not react with the reactive gas in the gas supply unit 11. In the gas supply unit 11, a reactive gas and a gas having a lower boiling point than the reactive gas react, so that a reactive cluster cannot be stably generated, and it may be difficult to process the sample. .

The generated reactive clusters are jetted from the jetting part 12 toward the sample 17 installed on the sample stage 14 by kinetic energy when jetted from the gas supply part 11. And by colliding with the surface of the sample 17, kinetic energy is converted into reaction energy and reacts with the sample.
In particular, by using a gas having high reactivity with the sample, for example, an interhalogen compound or hydrogen halide, the reaction between the reactive gas and the sample occurs in the collision portion of the reactive cluster, and atoms or molecules on the sample surface are removed. It can be etched and removed efficiently. For this reason, a sample can be processed efficiently by using the reactive gas and sample of a combination with high mutual reactivity on the sample surface.

  As described above, a sample can be processed without applying an accelerating voltage to the cluster by injecting a reactive gas and a reactive cluster of a gas having a lower boiling point than the reactive gas onto the sample. . Since this reactive cluster is not an ion, there is no need to neutralize the reactive cluster, and electrical damage to the sample that occurs when the cluster is not completely neutralized can be eliminated. .

In the above-described apparatus, the processing apparatus provided with one ejection part for ejecting the mixed gas has been described. However, in the processing method of the present invention, the processing apparatus may include a plurality of ejection parts. Good.
By processing the sample by ejecting clusters in the same direction from a plurality of ejection portions, it is possible to perform a wider range of processing than the processing using a single ejection portion, and the processing speed of the sample is improved.
Further, by providing a plurality of ejection portions having different ejection directions, it is possible to simultaneously perform cluster ejection from the vertical direction and cluster ejection from the oblique direction on the sample. In addition, by controlling the mixed gas pressure and flow rate at the plurality of jetting portions the same or individually, it is possible to control the surface condition and processing speed of the sample.
Furthermore, it is possible to combine a plurality of the above-described processing methods, and it becomes possible to perform a complicated shape faster and with a single processing as compared with the conventional method.

Next, an embodiment of the semiconductor element of the present invention that can be manufactured by using the above-described cluster injection processing method will be described. FIG. 17 shows a schematic configuration diagram of the semiconductor element of the present embodiment.
A semiconductor element 40 shown in FIG. 17 includes a base 51 and a semiconductor chip 41 mounted on the base 51.

The semiconductor chip 41 includes a multilayer wiring layer 50 formed of a wiring layer 43 and an interlayer insulating layer 44 formed on one main surface of a semiconductor substrate 42, and a connection electrode formed on the other main surface. A pad 49 and a through electrode 45 formed through the semiconductor substrate 42 are provided.
The base 51 is provided with a multilayer wiring layer 55 formed on one main surface of the base 51 and composed of a wiring layer 52 and an interlayer insulating layer 53, and an electrode pad 54 for connection.
The semiconductor chip 41 is mounted on the base 51 by connecting the electrode pads 54 formed on the base 51 and the electrode pads 49 formed on the semiconductor substrate 42.

The through electrode 45 formed in the semiconductor chip 41 is composed of a buried conductive layer 46 formed in the hole 56 formed by etching using the above-described reactive cluster.
Further, the through hole provided in the semiconductor substrate 42 is formed by the sample processing method using the reactive cluster described above.

  In the through electrode 45, an insulating layer 48 is formed on the side wall of the hole 56 in the hole 56 provided in the semiconductor substrate 42. A barrier metal layer 47 is formed on the insulating layer 48 along the side wall.

In addition, when the wiring layer 43 formed on the semiconductor chip 41 and the wiring layer 52 formed on the base 51 are formed by, for example, a subtractive method, they become wiring layers on the entire surface of the insulating layer. After the conductor layer is formed, a wiring pattern made of a photoresist or the like is formed, and the conductor layer is etched to form a wiring layer having a desired pattern.
For the etching of the conductor layer, the above-described sample processing method using reactive clusters can be used. Etching with the above-mentioned reactive cluster enables highly anisotropic etching, so there is no influence of thinning of the wiring due to side etching or undercut that occurs in the conventional subtractive method, and wiring design rules Can be miniaturized. Therefore, by using the above-described sample processing method using reactive clusters, high-density wiring is possible even when the subtractive method is used.

The semiconductor substrate 42 may have through holes formed by a method other than the above, but at least one or more through holes are formed by etching using the above-described reactive clusters. The wiring layer is preferably formed by etching using the reactive cluster, but may be etched by a method other than this method.
Therefore, in the above-described semiconductor element, any one or more of the hole provided in the semiconductor substrate or the wiring pattern may be formed by using the above-described reactive cluster processing. In addition to the processing by the reactive cluster, each of the above-described configurations formed by using other processing methods may be provided.

Next, an embodiment of a method for manufacturing a semiconductor device of the present invention will be described.
First, as shown in FIG. 18A, a wiring layer 43 and an interlayer insulating layer 44 are stacked on one main surface side of a semiconductor substrate 42 to form a multilayer wiring layer 50. Then, a hole 56 is formed on the semiconductor substrate 42 on which the multilayer wiring layer 50 is formed from the multilayer wiring 50 side. The hole 56 penetrates the multilayer wiring layer 50 and is formed partway in the thickness direction of the semiconductor substrate 42. The insulating layer 48 is formed in the formed hole using, for example, a thermal oxidation method or a CVD (Chemical Vapor Deposition) method.

The formation method of the above-mentioned hole part 56 applies the processing method using the above-mentioned reactive cluster.
First, for example, a pattern mask made of a photoresist that opens only a portion where the hole 56 is to be formed in the semiconductor substrate 42 is formed by photolithography, and the hole 56 is etched. For this etching, the above-described sample processing method using reactive clusters is used. For example, in the processing apparatus shown in FIG. 1, the semiconductor substrate 42 is placed on a sample stage, and a mixed gas of a reactive gas and a gas having a lower boiling point than the reactive gas is ejected from the ejection portion. Thereby, a reactive cluster is generated and the semiconductor substrate 42 is etched. When a silicon substrate is used as the semiconductor substrate 42, it is preferable to use a mixed gas of ClF ₃ gas having high reactivity with the silicon substrate and argon gas.

  For example, when forming the wiring layer 43 on the semiconductor substrate 42 by a subtractive method, a conductor layer to be a wiring layer is formed on the entire surface of the insulating layer, and a wiring pattern using a photoresist or the like is formed. The mask is formed. Then, the wiring layer 43 having a desired pattern is formed by etching the conductor layer using this mask.

  Etching with the above-described reactive clusters can be performed with high anisotropy. Therefore, deep etching with an aspect ratio of 1 or more can be performed precisely by etching using this reactive cluster, and a precise hole 56 can be formed in the semiconductor substrate 42. Further, by patterning the wiring layer 43 using a processing method using reactive clusters, the wiring layer 43 is not thinned by side etching or undercutting of the conductor layer, and the wiring design rules are made finer. In addition, high-density wiring is possible.

Next, as shown in FIG. 18B, a barrier metal layer 47 and a buried conductive layer 46 are formed in the hole 56.
First, the barrier metal layer 47 made of TiN, WN, Ti, TaN, Ta or the like is formed on the entire surface of the semiconductor substrate 42 including the hole 56 by using a sputtering method or a CVD method. Then, a conductor layer made of Cu, Al, Ti, Sn or the like is formed on the barrier metal layer 47. Then, by using a CMP (Chemical Mechanical Polishing) method or the like, the conductor layer and the barrier metal layer 47 formed on the semiconductor substrate 42 excluding the inside of the hole 46 are removed, whereby the barrier 56 is formed in the hole 56. A metal layer 47 and a buried conductive layer 46 are formed.

  Further, the substrate may be planarized by a processing method using the above-described reactive cluster instead of the CMP method. For example, in the processing apparatus shown in FIG. 1, the semiconductor substrate 42 is placed on a sample stage, and a mixed gas of a reactive gas and a gas having a lower boiling point than the reactive gas is ejected from the ejection portion. At this time, by moving the sample stage, reactive clusters can be sprayed over the entire surface of the sample, and the surface of the sample can be flattened. Thereby, a reactive cluster is generated, and the conductor layer and the barrier metal layer 47 formed on the semiconductor substrate 42 are etched. As the reactive gas, it is preferable to use a gas such as HCl, HBr, or HI that has high reactivity with the metal used for the conductor layer or the barrier metal layer 47.

Next, as shown in FIG. 18C, the embedded conductive layer 46 is exposed from both main surfaces of the semiconductor substrate 42 to form a through electrode 45, and a wiring layer 43 and an electrode pad 49 connected to the through electrode 45. Form.
First, the surface of the semiconductor substrate 42 where the multilayer wiring layer 50 is not formed is etched by CMP or etching using the above-described reactive cluster. By this etching, the buried conductive layer 46 is exposed from the semiconductor substrate 42. In this manner, the embedded conductive layer 46 is exposed from both main surfaces of the semiconductor substrate 42 to form the through electrode 45.
Then, the wiring layer 43 connected to the through electrode 45 is formed on the multilayer wiring layer 50 by the same method as the wiring layer 43 of the multilayer wiring layer 50 described above. An electrode pad 49 for connecting the through electrode 45 is formed on the other surface.

The semiconductor chip 41 can be formed by the above process. Then, by connecting the electrode pad 49 of the semiconductor chip 41 to the electrode pad 54 of the base 51, the semiconductor element 40 shown in FIG. 17 can be manufactured.
The base 51 can be manufactured by the same method as the above-described method for manufacturing the semiconductor chip 41 except for the step of forming the hole 56 and the through electrode 45, and can be manufactured by using other known methods. be able to.

Next, an embodiment of a micro electro mechanical system (MEMS) of the present invention that can be manufactured using the above-described cluster jet processing method will be described.
In FIG. 19, the schematic block diagram of a pressure sensor is shown as embodiment of MEMS.
A pressure sensor 20 shown in FIG. 19 has a configuration in which a controller-integrated MEMS sensor is formed on a semiconductor substrate 21.

  On one main surface of the semiconductor substrate 21, a MEMS sensor controller 22 composed of a MOS (Metal Oxide Semiconductor) transistor or a MOS capacitor is formed. Further, a low-k wiring 23 made of a low dielectric constant material (low-k) connected to the MEMS sensor controller 22 is formed. An insulating passivation film 24 is formed on the semiconductor substrate 21 and the low-k wiring 23.

  A vacuum cover glass 28 is attached to the surface of the semiconductor substrate 21 where the MEMS sensor / controller 22 is formed. The vacuum cover glass 28 is provided with a capacitor / controller wiring 29 on the side to be bonded to the semiconductor substrate 21. The capacitor / controller wiring 29 is connected to the MEMS sensor / controller 22 via a low-k wiring 23. The vacuum cover glass 28 is bonded onto the passivation film 24 of the semiconductor substrate 21 with an adhesive layer 31.

  In addition, a spin-on-glass (SOG) insulating layer 25 is formed on the other main surface of the semiconductor substrate 21. An electrode layer 26 made of Ni or the like is formed on the SOG insulating layer 25.

  The above-described semiconductor substrate 21 is formed with a pressure reference vacuum 27 composed of holes having an aspect ratio of 1 or more that penetrate both main surfaces. The pressure reference vacuum 27 of the semiconductor substrate 21 is formed by the sample processing method using the above-described reactive cluster. Then, by forming the pressure reference vacuum 27 on the semiconductor substrate 21, the membrane structure 30 is formed in the pressure sensor 20 by the SOG insulating layer 25 and the electrode layer 26 at a portion in contact with the pressure reference vacuum 27. The capacitor / controller wiring 29 of the vacuum cover glass 28 is formed at a position facing the membrane structure 30 with the semiconductor substrate 21 interposed therebetween.

Next, a method for manufacturing the pressure sensor 20 will be described.
First, the MEMS sensor controller 22 is formed on one main surface side of the semiconductor substrate 21. The MEMS sensor controller 22 can be manufactured by the same method as a known MOS transistor and MOS capacitor. Further, a low-k wiring 23 for the MEMS sensor controller 22 is formed on the MEMS sensor controller 22 with a low-k material. Then, a passivation film 24 that covers the semiconductor substrate 21, the MEMS sensor / controller 22, and the low-k wiring 23 is formed.
Further, the SOG insulating layer 25 is formed on the other main surface side of the semiconductor substrate 21 by using a spin coat method or the like. Then, the electrode layer 26 is formed on the SOG insulating layer 25 by, for example, Ni plating.

Next, as shown in FIG. 20, a pattern mask made of a photoresist 32 that opens only a portion where the pressure reference vacuum 27 is formed in the semiconductor substrate 21 is formed by photolithography.
Then, the semiconductor substrate 21 is etched using the photoresist 32 as a mask. For this etching, the above-described sample processing method using reactive clusters is used. For example, in the processing apparatus shown in FIG. 1, the semiconductor substrate 21 is placed on a sample stage, and a mixed gas of a reactive gas and a gas having a boiling point lower than that of the reactive gas is ejected from an ejection part. Thereby, a reactive cluster is generated and the semiconductor substrate 21 is etched. When a silicon substrate is used as the semiconductor substrate 21, it is preferable to use a mixed gas of ClF ₃ gas and argon gas, which has high reactivity with the silicon substrate.
By etching using this reactive cluster, deep digging with an aspect ratio of 1 or more can be performed only on the portion of the semiconductor substrate 21 where the pressure reference vacuum 27 is formed.

Further, after forming a hole serving as a pressure reference vacuum 27, a vacuum cover glass is pasted onto the passivation film 24 of the semiconductor substrate 21 using the adhesive layer 31 in the etched vacuum processing chamber. Capacitor and controller wiring 29 is formed in advance on the vacuum cover glass. Then, a vacuum cover glass 28 is attached to the semiconductor substrate 21 so that the capacitor / controller wiring 29 is connected to the low-k wiring 23.
Through the above steps, a capacitance pressure sensor integrated with the controller shown in FIG. 19 can be manufactured.

  According to the manufacturing method described above, since the reactive cluster is used in the etching process of the semiconductor substrate 21, plasma anisotropic etching is not performed. For this reason, even when the MEMS sensor controller 22 that is generally susceptible to plasma damage is formed on the semiconductor substrate 21 before the etching process, the characteristics are not affected by the threshold voltage due to plasma heat or the design value. In addition, the change in dielectric constant of the low-k interlayer insulating film due to plasma heat can be eliminated. Therefore, a controller-integrated MEMS sensor with little deviation from the design value can be manufactured on the substrate.

  Next, an embodiment of the optical component of the present invention that can be manufactured using the above-described cluster injection processing method will be described. FIG. 21A shows a manufacturing process diagram of an exposure mask, and FIG. 21B shows a schematic configuration diagram of an exposure mask manufactured by the processing method using the above-described reactive cluster.

  As shown in FIG. 21A, a metal film 64 made of Mo, Ta, Cr or the like serving as a light shielding layer is formed on the entire surface of the quartz substrate 61. Then, a photoresist pattern is formed on the metal film 64 only on the metal film 64 where the light shielding layer is to be formed, for example, by photolithography.

  Next, using the formed photoresist as a mask, the metal film 64 is etched using the above-described sample processing method using reactive clusters. For example, in the processing apparatus shown in FIG. 1, a quartz substrate 61 is placed on a sample stage, and a mixed gas of a reactive gas and a gas having a boiling point lower than that of the reactive gas is ejected from an ejection part. Thereby, a reactive cluster is generated and the metal film 64 on the quartz substrate 61 is etched. As the reactive gas, it is preferable to use a gas such as HCl, HBr, or HI that has high reactivity with the metal used for the metal film 64.

  After the metal film 64 is etched using the reactive cluster, the photoresist is removed. Through the above steps, as shown in FIG. 21B, an exposure mask having an optical pattern made of the light shielding layer 62 can be manufactured on the quartz substrate 61.

  Next, an optical slit will be described as another embodiment of the optical component of the present invention. FIG. 22A shows a manufacturing process diagram of the optical slit of the present embodiment, and FIG. 22B shows a schematic configuration diagram of the optical slit manufactured by a processing method using reactive clusters.

  As shown in FIG. 22A, a photoresist pattern is formed on the silicon substrate 71 by, for example, photolithography, except for a portion where a hole for forming a slit is formed.

Next, using the formed photoresist as a mask, the silicon substrate 71 is etched using the above-described sample processing method using reactive clusters. For example, in the processing apparatus shown in FIG. 1, a silicon substrate 71 is placed on a sample stage, and a mixed gas of a reactive gas and a gas having a boiling point lower than that of the reactive gas is ejected from an ejection part. Thereby, a reactive cluster is generated and the silicon substrate 71 is etched. For etching the silicon substrate 71, it is preferable to use a mixed gas of ClF ₃ gas and argon gas, which has high reactivity with the silicon substrate 71.
Then, the silicon substrate 71 is etched using reactivity, and then the photoresist is removed.
Through the above steps, the hole 72 can be formed in the silicon substrate 71 as shown in FIG. 22B. Moreover, the slit 73 can be formed from the hole 72 formed by etching and the remaining silicon substrate 71.
Thus, the optical slit 70 provided with the optical pattern which consists of the slits 73 can be manufactured by the above-described manufacturing method as shown in FIG. 22B.

  In an optical component such as the above-described exposure mask and optical slit, an optical pattern having a high anisotropy can be etched by forming an optical pattern such as a light shielding layer or an optical slit by a reactive cluster. For this reason, the etching using the reactive cluster does not cause a decrease in pattern accuracy due to side etching or undercut during the formation of the optical pattern, and the optical pattern can be miniaturized and densified. .

Next, the result of actually processing a sample using the above-described processing method by cluster injection according to the present embodiment will be described below.
First, the sample surface was processed by changing the injection time of the mixed gas injected to the sample and the primary pressure of the mixed gas in the gas supply unit. FIG. 2 shows the relationship between the primary pressure by this processing, the injection time to the sample, and the etching rate of the sample.

As processing conditions, a silicon single crystal having a natural oxide film with a thickness of about 2 nm formed on the surface is used as a sample, ClF ₃ gas 6 vol% as a reactive gas, and Ar gas 94 vol as a gas having a lower boiling point than the reactive gas. % Mixed gas was used. Moreover, the diameter of the ejection part was 0.1 mm. The primary pressure was changed as shown in FIG. 2, and the injection time was changed to 2, 5, 7, and 15 minutes.

  From the results shown in FIG. 2, it can be seen that the etching rate of the sample is increased by increasing the primary pressure under the conditions of the same injection time. Therefore, by increasing the primary pressure and increasing the difference from the secondary pressure, the kinetic energy of the generated reactive clusters can be increased, and the reaction energy generated by the collision between the reactive clusters and the sample can be increased. can do. For this reason, the etching rate can be increased.

  Next, in the apparatus shown in FIG. 1, the sample surface was processed by changing the distance between the ejection part and the sample surface to 26 mm, 35 mm, and 50 mm. FIG. 3 shows the relationship between the distance between the ejection portion and the sample surface and the etching rate of the sample by processing. Moreover, the photograph showing the state of the sample surface after processing the distance between the ejection part and the sample surface to 26 mm is shown in FIG. 4, and the state of the sample surface after processing the distance between the ejection part and the sample surface to 35 mm is shown. The photograph to represent is shown in FIG.

The processing conditions of the sample were a silicon single crystal as a sample, and a mixed gas composed of 6 vol% ClF ₃ gas as a reactive gas and 94 vol% Ar gas as a gas having a lower boiling point than the reactive gas. The reactive cluster injection time to the sample was 5 minutes, the flow rate of the mixed gas was 398 sccm, the primary pressure was 0.53 MPaG, the secondary pressure was 3.2 Pa, and the temperature in the vacuum processing chamber was 28 ° C.

As shown in FIG. 3, when the distance between the ejection part and the sample was 26 mm, an etching rate of about 4.1 μm / min was obtained. In addition, when the distance between the ejection part and the sample was 35 mm, an etching rate of about 2.2 μm / min was obtained. Further, the etching rate became almost zero when the distance was 50 mm or more.
In addition, when the distance between the ejection part and the sample was 26 mm, an etching rate approximately twice as high as that when the distance was 35 mm could be obtained.

From the above results, it can be seen that the etching rate of the sample increases as the distance between the ejection portion and the sample is reduced.
In addition, since the etching rate of the sample changes depending on the distance between the ejection part and the sample, the etching rate of the sample can be changed by changing the distance between the ejection part and the sample.

  Next, the sample surface was processed by tilting the sample stage so that the incident angle of the cluster was oblique to the sample surface. Further, a silicon single crystal was used as a sample, and a line-and-space pattern was formed on the surface by a photoresist layer having a thickness of 1.2 μm. A photograph of a cross section near the processed sample surface is shown in FIG.

The sample processing conditions were such that the distance between the ejection part and the sample was 26 mm, and a mixed gas composed of 6 vol% ClF ₃ gas as the reactive gas and 94 vol% Ar gas as the gas having a lower boiling point than the reactive gas. The reactive cluster injection time to the sample was 5 minutes, the flow rate of the mixed gas was 398 sccm, the primary pressure was 0.53 MPaG, the secondary pressure was 3.2 Pa, and the temperature in the vacuum processing chamber was 28 ° C.

As shown in FIG. 6, the sample is etched obliquely in accordance with the incident direction of the reactive cluster. Thus, by performing incidence from an oblique direction, etching with high straightness corresponding to the incident angle is possible.
In addition, as described above, by utilizing the fact that the etching rate of the sample changes due to the change in the distance between the ejection part and the sample, a state in which the distance between the sample and the nozzle is kept constant during processing from an oblique direction Then, the ejection part or the sample is moved. Thereby, a projection part is etched preferentially. Therefore, it is possible to reduce the level difference on the sample surface and flatten it.

  Further, as shown in FIG. 6, the photoresist layer formed on the sample surface was hardly etched, and the etching selectivity between the silicon single crystal and the photoresist was 100: 1 or more. For this reason, after forming a pattern using the mask of a photoresist, a silicon single crystal can be selectively processed and a pattern can be formed.

Next, the sample was penetrated by injecting reactive clusters into the sample.
Through holes were formed by injection of reactive clusters. A photograph of the state of the silicon single crystal in which the through hole is formed is shown in FIG.

The processing conditions of the sample are as follows: the distance between the ejection part and the sample is 26 mm, a silicon single crystal having a thickness of 625 μm is used as the sample, ClF ₃ gas is 6 vol% as the reactive gas, and the gas has a lower boiling point than the reactive gas. A mixed gas consisting of 94 vol% Ar gas was used. In addition, the injection time of the reactive cluster to the sample was 15 minutes, the flow rate of the mixed gas was 585 sccm, the primary pressure was 0.8 MPaG, the secondary pressure was 7.5 Pa, and the temperature in the vacuum processing chamber was 25.4 ° C. .

  As shown in FIG. 7, through holes could be formed in a silicon single crystal having a thickness of 625 μm by the above-described processing. Moreover, since a through-hole was able to be formed in a silicon single crystal having a thickness of 625 μm within an injection time of 15 minutes or less, the etching rate under the above-described processing conditions is 41 μm / min or more.

  From the above processing results, a through hole can be formed in the sample by injecting the reactive gas and a reactive cluster of a gas having a lower boiling point than the reactive gas onto the sample. For this reason, for example, it becomes possible to form a scribe line in a silicon material, or to cut a substrate at the formed scribe line. Furthermore, using a good selection ratio with the photoresist, an arbitrary pattern is formed on the sample by using a photoresist mask, a scribe line other than a straight line is formed, or the substrate is cut at the formed scribe line. It becomes possible to do.

Next, instead of using the mixed gas in the gas supply unit, the sample surface was processed using ClF ₃ gas alone. A photograph showing the state of the processed sample surface is shown in FIG.

  The processing conditions of the sample were: the distance between the ejection part and the sample was 26 mm, the injection time of the reactive cluster to the sample was 15 minutes, the flow rate of the mixed gas was 300 sccm, the primary pressure was 0.05 MPaG, and the secondary pressure Was 2.2 Pa, and the temperature in the vacuum processing chamber was 29 ° C.

As shown in FIG. 8, although a trace of processing by reactive clusters is formed on the sample surface, the sample surface can hardly be etched with respect to the spray time of 15 minutes, and the etching rate under the above conditions is 0.01 μm / min or less.
Thus, ClF ₃ gas has a high boiling point, want for liquefied under high pressure, in the case of using a ClF ₃ gas alone, in the processing of the sample can not be increased primary pressure, secondary The difference with pressure cannot be increased. Therefore, sufficient adiabatic expansion for generating reactive clusters cannot be obtained.
Therefore, by mixing the reactive gas with a gas having a lower boiling point than the reactive gas and increasing the primary pressure, a cluster of mixed gas injected from the injection unit is sufficiently generated, and the sample surface is processed efficiently. It was proved.

  Next, the sample surface was processed using Ar gas alone instead of the mixed gas. A photograph showing the state of the processed sample surface is shown in FIG.

  The sample processing conditions were such that the distance between the ejection part and the sample was 26 mm. The reactive cluster injection time to the sample was 15 minutes, the Ar gas flow rate was 637 sccm, the primary pressure was 0.8 MPaG, the secondary pressure was 8.6 Pa, and the temperature in the vacuum processing chamber was 21.1 ° C. .

As shown in FIG. 9, in the processing using Ar gas, no change was observed on the sample surface. For this reason, it turns out that the processing using a single Ar gas cannot process the silicon single crystal.
Therefore, it was proved that the sample surface could be processed by using a reactive gas as the mixed gas and a gas having a lower boiling point than the reactive gas.

  Next, the diameter of the ejection part was enlarged to 6 mm, and the sample surface was processed with a primary pressure of -0.09 MPaG. A photograph showing the state of the sample surface after processing is shown in FIG.

The sample processing conditions were such that the distance between the ejection part and the sample was 26 mm, and a mixed gas composed of 6 vol% ClF ₃ gas as the reactive gas and 94 vol% Ar gas as the gas having a lower boiling point than the reactive gas. Further, the injection time of the mixed gas to the sample was 15 minutes, the flow rate of the mixed gas was 580 sccm, the secondary pressure was 7.5 Pa, and the temperature in the vacuum processing chamber was 25.4 ° C.

Under the above-mentioned conditions, the diameter of the ejection part was enlarged to 6 mm, the primary pressure was lowered, and the difference from the secondary pressure was reduced. As a result, as shown in FIG. 10, no processing trace was formed on the sample surface.
From this result, it is considered that the cluster of the mixed gas was not effectively formed under the above-described processing conditions because the primary pressure was too low. From this result, it can be said that a differential pressure between the primary pressure and the secondary pressure is necessary for the generation of the reactive cluster.

Next, the sample surface was processed using Cl ₂ gas or F ₂ gas instead of the interhalogen compound or hydrogen halide used as the reaction gas. A photograph showing the state of the sample surface processed using Cl ₂ gas is shown in FIG. 11, and a photograph showing the state of the sample surface processed using F ₂ gas is shown in FIG.

The processing conditions in the case of using Cl ₂ gas were such that the distance between the ejection part and the sample was 26 mm, and a mixed gas composed of 5 vol% of Cl ₂ gas and 95 vol% of Ar gas as the inert gas was used. Further, the injection time of the mixed gas to the sample was 15 minutes, the flow rate of the mixed gas was 542 sccm, the primary pressure was 0.81 MPaG, the secondary pressure was 8.3 Pa, and the temperature in the vacuum processing chamber was 28.4 ° C.

The processing conditions when F ₂ gas is used include a distance between the ejection part and the sample of 26 mm, F ₂ gas as a mixed gas, ₅ vol%, and Ar gas as a boiling point lower than reactive gas, 95 vol%. A mixed gas was used. Further, the injection time of the mixed gas to the sample was 30 minutes, the flow rate of the mixed gas was 569 sccm, the primary pressure was 0.81 MPaG, the secondary pressure was 8.6 Pa, and the temperature in the vacuum processing chamber was 27.4 ° C.

As shown in FIG. 11 and FIG. 12, although the portion where the mixed gas was injected changed color, the sample surface was hardly etched. From this result, it is understood that the sample cannot be processed when a mixed gas of halogen and inert gas is used.
From this result, it is understood that the sample surface can be processed by using a mixed gas of a reactive gas such as an interhalogen compound and hydrogen halide and a gas having a lower boiling point than the reactive gas.

  Next, the sample surface was processed using a sample in which a silicon oxide film was formed on the surface of a silicon single crystal. A photograph showing the state of the sample surface after processing is shown in FIG.

The sample processing conditions were such that the distance between the ejection part and the sample was 26 mm, and a mixed gas composed of 6 vol% ClF ₃ gas as the reactive gas and 94 vol% Ar gas as the gas having a lower boiling point than the reactive gas. Further, the injection time of the mixed gas to the sample was 15 minutes, the flow rate of the mixed gas was 560 sccm, the primary pressure was 0.7 MPaG, the secondary pressure was 5 Pa, and the temperature in the vacuum processing chamber was 28.2 ° C.

  As shown in FIG. 13, the silicon oxide film formed on the sample surface is not etched under the above processing conditions. Therefore, it can be seen that, in the above-described processing method, the silicon oxide film has a very low etching rate compared to the silicon single crystal, and the etching selectivity between the silicon oxide film and the silicon material is large. For this reason, in addition to forming a pattern using a photomask, a mask pattern made of a silicon oxide film, a silicon nitride film, a silicon oxynitride film, or the like is formed on the surface of the silicon material. Processing becomes possible.

Thus, by forming a film resistant to reactive gas on the surface of the sample as a pattern mask, the sample can be processed according to the pattern.
In addition to the materials described above, a metal material can also be used as the material for the pattern mask. When the reactive gas is an interhalogen compound, it is not necessary to consider the reactivity with the metal material. Therefore, a sample using a metal material mask by applying a metal material used as a general pattern mask. Can be processed.
When the reactive gas is hydrogen halide, a metal material such as Ni that is resistant to hydrogen halide is selected, and the sample can be processed using a mask made of the metal material.

  Next, a natural oxide film having a thickness of about 2 nm formed on the surface of a silicon single crystal serving as a sample was removed by treatment with hydrogen fluoride (HF), and then the sample surface was processed. A photograph showing the state of the surface after processing is shown in FIG. For comparison, FIG. 15 shows a photograph showing the state of the surface after processing for a sample that has been subjected to surface processing without removing the natural oxide film.

The sample processing conditions were such that the distance between the ejection part and the sample was 26 mm, and a mixed gas composed of 6 vol% ClF ₃ gas as the reactive gas and 94 vol% Ar gas as the gas having a lower boiling point than the reactive gas. Further, the injection time of the mixed gas to the sample was 5 minutes, the flow rate of the mixed gas was 560 sccm, the primary pressure was 0.3 MPaG, the secondary pressure was 1.5 Pa, and the temperature in the vacuum processing chamber was 28.2 ° C.
Further, for comparison, a sample which was surface-treated without removing the natural oxide film was shown in FIG. 2 as ClF ₃ gas 6 vol% as a reactive gas and Ar gas 94 vol as a gas having a lower boiling point than the reactive gas. % Is a sample processed using a mixed gas consisting of 1%, a primary pressure of 0.3 MPaG, and an injection time of 5 minutes.

  Comparing the result shown in FIG. 14 with the result shown in FIG. 15, the etching depth of the sample from which the natural oxide film shown in FIG. 14 was removed was 8.67 μm, whereas that shown in FIG. The etching depth of the sample from which the natural oxide film was not removed was 0.711 μm.

  The etching rate of the sample from which the natural oxide film has not been removed is 0.14 μm / min, whereas the etching rate of the sample from which the natural oxide film has been removed is 1.73 μm / min. As a result, the etching rate was increased about 12 times by removing the natural oxide film formed on the surface of the silicon single crystal with HF or the like.

FIG. 16 shows the relationship between the primary pressure by this processing, the injection time to the sample, and the etching rate of the sample. In FIG. 16, for comparison, the relationship between the injection time to the sample shown in FIG. 2 and the etching rate of the sample is described as it is, and the processing of the sample performed by removing the natural oxide film described above is an asterisk. And arrows.
As shown in FIG. 16, the etching rate is greatly improved by removing the natural oxide film as compared with the case where the sample from which the natural oxide film has not been removed is processed under substantially the same primary pressure and injection time. I understand that.
Therefore, the etching rate of the sample can be improved by removing the natural oxide film of the silicon material.

  The present invention is not limited to the above-described configuration, and various other configurations can be employed without departing from the gist of the present invention.

DESCRIPTION OF SYMBOLS 11 Gas supply part 12 Ejection part 13 Vacuum processing chamber 14 Sample stand 15 Turbo molecular pump 16 Dry pump 17 Sample 18 Molten gas molecule 19 Reactive cluster 20 Pressure sensor 21 Semiconductor substrate 22 MEMS sensor controller 23 Low-k wiring 24 Passivation Membrane 25 SOG insulating layer 26 Electrode layer 27 Pressure reference vacuum 28 Cover glass 29 Capacitor / controller wiring 30 Membrane structure 31 Adhesive layer 32, 63, 74 Photoresist 40 Semiconductor element 41 Semiconductor chip 42 Semiconductor substrate 43, 52 Wiring layer 44 , 53 Interlayer insulating layer 45 Through electrode 46 Embedded conductive layer 47 Barrier metal layer 48 Insulating layer 49, 54 Electrode pad 50, 55 Multilayer wiring layer 51 Base 56, 72 Hole 60 Exposure mask 61 Quartz substrate 62 light shielding layer 64 metal film 70 optical slit 71 silicon substrate 73 slit

Claims (11)

A gas mixture consisting of a reactive gas and a gas having a boiling point lower than that of the reactive gas is ejected while adiabatically expanding in a predetermined direction from the ejection section at a pressure that does not liquefy. A cluster injection processing method comprising: generating and injecting the reactive cluster onto a sample in a vacuum processing chamber and processing the sample surface by a chemical reaction.   The cluster injection processing method according to claim 1, wherein the reactive gas is an interhalogen compound gas or a hydrogen halide gas.   The cluster injection processing method according to claim 1, wherein the sample is a semiconductor material or a metal material.   Forming a pattern on the surface of the sample using one or more of a photoresist, a silicon oxide film, a silicon nitride film, or a silicon oxynitride film, and processing the sample using the pattern as a mask. The cluster injection processing method according to claim 1, wherein:   2. The cluster jet processing method according to claim 1, wherein a pattern is formed on the sample surface using a metal material resistant to the reactive gas, and the sample is processed using the pattern as a mask. .   The cluster injection processing method according to claim 1, wherein the sample is processed obliquely by injecting the reactive cluster with respect to the sample surface from an oblique direction.   The cluster injection processing method according to claim 6, wherein the sample surface is flattened by moving the ejection part or the sample. The cluster injection processing method according to claim 1, wherein linear processing, curved processing, and wide area processing are performed by moving the ejection portion or the sample.   The cluster injection processing method according to claim 1, wherein a through hole is formed in the sample.   2. The cluster jet processing method according to claim 1, wherein the surface is processed after removing the natural oxide film on the sample surface. A processing method in which the reactive gas is ejected from the plurality of ejection portions in the same direction to eject reactive clusters to the sample, and the reactive gas is ejected from the plurality of ejection portions in different directions to react with the sample. One type or two or more types of processing selected from a processing method for jetting reactive clusters, or a processing method for jetting reactive clusters to the sample by controlling the flow rate and pressure of a plurality of jetting portions either the same or individually The cluster injection processing method according to claim 1, wherein the methods are performed simultaneously.